Theory of Polarographic Currents Controlled by Rate of Reaction and

Polarographic Study of Several cis and trans Coördination Compounds of Cobalt. Henry F. Holtzclaw Jr. , David P. Sheetz. Journal of the American Chem...
0 downloads 0 Views 708KB Size
404'1 system ethanol-water (Table 111), equation 11 is valid to an excellent approximation, and for @ = 0.628 the probable error of the fit is only 0.002. p values for the other solvents have been used to estimate Y- with the aid of equation 11. These tentative estimates are listed in Table 111. For the solvent butanol-1, Y- was also computed from data for aliphatic acids using equation 8. The value obtained in this way is in satisfactory agreement with the one based on equation 11 (see Table 111), so that the present correlations do not seem to be limited in scope to the system ethanol-water. TABLE 111 On the Determination of Degenerate Single Ion REACTION CONSTANTS FOR THE IONIZATION OF SUBSTITUTED Activity Coefficients.-If equation A is accepted, BENZOICACIDSIS VARIOUSS O L V E N ~22 S , f 4" the way is open for the determination of log Solvent li YReferenced f~ values and thus of other degenerate single ion Water 1 000 0 000 41 activity coefficients. From the data reported in 50 Vol. % ethanol 1.464 0 737" 18 this article alone it is not possible to solve equations Ethanol 1 626 1 000 17 6 for log fH since there is one more unknown than 50 Vol. 70methanol 1 241 0 3846 42 there are equations. However when values of ~ K A Methanol 1 374 0 595b 17 and of the activity function for other structural Butanol-1, 0 0.5 ili LiCl 1 476 0 758b'C 15 types become available, the number of independent a By interpolation from Table 11. Calculated from equations increases more rapidly than the number equation 11. Using the data of Wooten and HammettI6 of new unknowns, and solutions for ??%A and log f H for aliphatic acids and benzoic acid in butanol-I, and of Table I1 for the same acids in 50.1% ethanol, the ratio of Y- become possible. This phase of the problem is values for the two solvents has been computed from equa- being actively pursued. tion 8. On this basis, Y- = 0.727 for butanol-l. In some Acknowledgments.-It is a pleasure to thank cases, p values were computed by the present authors from Professor L. P. Hammett in whose laboratory the published data. much of this work was done for his great hosIt follows from equations A and 10 that the vari- pitality, and t o thank him and Professor D. Y. ation of p with solvent is given by equation 11 Curtin for many valuable discussions. One of where 6 is a parameter On the basis of data for the us (E. G.) is grateful to the Frank B. Jewett Fellowship Committee for the award of a fellow._P = PL @Y(11) ship which has made this work possible. (12) R Kuhn and . 4 XTasserniann, H F ~ Z ( / ~ ~ iInd a 11, 1 31, i t represented by equation 10. In this equation, K and KOare equilibriuin or rate constants for the substituted and unsubstituted materials, u is the log K = log K O p c (10) substituent function for which Hammett has tabul a t e ~ empirical l~~ values, and p is the reaction constant. For the ionization of substituted benzoic acids, the reaction constant in water, pw, is equal to unity, and p varies greatly with solvent (see Table TII).

+

~

+

TA1,I2AH.4SSEE, F I , A ,

(19281

[CONTRIBUTION FROM

THE

RECEIVEDMARCH13, 1950

COATES CHEMICAL LABORATORY, LOUISIANA STATE UNIVERSITY]

Theory of Polarographic Currents Controlled by Rate of Reaction and by Diffusion1 BY PAULDELARAY An equation is derived for polarographic currents controlled by rate of reaction and by diffusion. The Ilkovic equation corresponds t o a special case of the more general equation reported in the present paper. The average limiting current is calculated. Variations of the limiting current with the head of mercury are discussed quantitatively. A simple graphic method for the computation of rate constants from experimental data is reported. The theory is applied t o the reduction of weak acids. Experimental data confirming the theoretical conclusions are presented for pyruvic acid.

Two quantitative treatments of polarographic currents controlled by rate of reaction and by diffusion have been reported in the recent years. The theory of Brdicka and Wiesner,2 although interesting, is based on rather arbitrary hypotheses as pointed out by Lir~gane.~The more rigorous treatment of Koutecky and Brdicka4 involves elaborate mathematical operations. A new treatment is discussed in the present paper and some of the difficulties of previous theories are eliminated. New (1) Paper presented before the division of Physical and Inorganic Chemistry of the International Congress of Pure and Applied Chemistry held in New York in September, 1951. (2) R. Brdicka and R. Wiesner, CoElcclion Csechoslow. Chcm. Commirn., 12, 138 (1947). (3) J. J. Lingane, Anal Chcm., 21, 45 (1949). (4) J. Koutecky and R Brdicka, Collecfion Czcchoslou. Chcm Commun , 12, 337 (1947)

features of diffusion-rate controlled currents are also reported. Case of Linear Diffusion Boundary Condition.-We consider the electrolytic reduction of two substances B and R which are transformed into one another according to the reaction B irs, R. We assume that R is reduced a t less negative potentials than B, and that the concentration of R is negligible in comparison with that of B. iMoreover, the potential of the electrode is adjusted in such a manner that only R is reduced. Under these conditions, B is transformed into R in the immediate neighborhood of the electrode as the reduction of R proceeds. If the transformation B irs, R is a first order reaction the number of moles B transformed into R a t the surface of the electrode

Oct., 1951

POLAROGRAPHIC CURRENTCONTROLLED BY REACTION RATEAND DIFFUSION

is proportional to the concentration C of B a t the surface of the electrode. Thus one has for an area equal to 1 sq. cm. ~ N= BkCdt

(1)

where k is the rate constant in cm.sec.-l, and Cis in moles The backward reaction R --+ B is neglected in equation 1 because the concentration of R a t the electrode is negligible since R is reduced as soon as i t is formed. The rate constant k in equation (1)is expressed in cm. see.-’ whereas rate constants of first order homogeneous reactions are in see.-’. A conventional rate constant k‘ can be introduced by writing equation 1 in the following manner di% = k’CGdt

(2)

where 6 represents the thickness (in cm.) of a monolayer of reducible substance a t the surface of the electrode. The product C6 is thus the surface concentration of reducible substance in moles cm. -2. As a first approximation, one may assume that 6 represents the average distance between two ions or molecules of the reducible substance in solution. The distance 6 is thus easily calculable on the basis of the concentration of reducible substance. However, the introduction of a surface concentration C6 is somewhat artificial and consequently we shall apply formula 1 rather than 2. The number dNB of moles of B diffusing toward the electrode is proportional to the gradient of concentration at the surface of the electrode and to the diffusion coefficient D of substance B. Thus one has for an area equal to 1 sq.

where x is the distance from the electrode. Combining equations (1) and (3), one obtains the following boundary condition

Derivation of the Current.-Concentration C of substance B is the solution of the equation for linear diffusion and for the boundary condition corresponding to equation (4). An identical problem is studied in the theory of heat conduction and the solution reported in the literature6 is applicable to the present case. The equation for C thus obtained has been applied by Delahay and Jaff6’ in their study of the kinetics of heterogeneous processes controlled by rate of reaction and by diffusion, The current i flowing through the electrolytic cell is proportional to the flux of substance a t the electrode, to the charge exchanged per mole of substance reduced, and to the area of the electrode.s The flux a t the electrode is calculated as follows. The equation giving the concentration as a function of x and t is differentiated with respect to x . The value x = 0 is introduced in the derivative dC/ ( 5 ) I . M . Kolthoff and J. J. Lingane, “Polarography,” Interscience Publishers, I n c , New York, N . Y., 1941, p. 17. (6) P . Frank and R . von Misses. “Die Differential- und Integralgleichungen der Mechanik und Physik,” Vol. IT, Rosenberg, New York, N.Y., 1936, p. 579. (7) P . Delabay and G. Jaff‘e, unpublished investigation. (8) See ref. (51, p. 21.

4945

a x , and the resulting equation is multiplied b y

D.

The equation thus obtained represents the flux a t the electrode. By multiplying the flux by the charge involved in the electrode process and by the area of the electrode, one obtains the current. Thus i

=

nFACok [l

- @(k&)]

exp 7 kat

(5)

where n is the number of electrons involved in the electrode process, F is the faraday, A is the area of the electrode in sq. cm., Co is the concentration of rek is the rate conacting substance in moles stant in cm. sec.-l, D is the diffusion coefficient of the reacting substance in see.-’, t is the time elapsed since the beginning of the electrolysis in sec., and represents the error integral

+(k.\/m

Equation (5) shows that the current decreases in the course of time. At time t = 0 the current has a finite value equal to nFACok, in contrast with the value i = m for a current entirely controlled by diffusion. By introducing the value k = 03 in formula (5), one should obtain the equation for the case of a current entirely controlled by diffusion. It is easily shown that this is indeed the case by expanding the error integral into a semi-convergent seriesg in the equation thus and by introducing k = obtained. Case of the Dropping Mercury Electrode In principle the method of calculation discussed in the previous section could be applied to the case of the dropping mercury electrode. However, the mathematical treatment becomes exceedingly intricate and i t is much simpler to apply the following method. The area of the electrode in formula ( 5 ) is replaced by its value calculated in terms of the characteristics m and t of the dropping mercury electrode. Moreover, the error integral is exis introduced in the panded and the value k = equation thus obtained. The resulting formula is identical with the Ilkovic equation except for the numerical constant. The correct equation is obtained by multiplying by .\/@ the second member of the formula thus derived from ( 5 ) . This is precisely the numerical coefficient which was used by Ilkoviclo to adjust the results obtained by application of the equation of linear diffusion to the case of the dropping mercury electrode. Consequently, one obtains the equation for a finite value of k by multiplying the second member of formula ( 5 ) by and by introducing in the resulting equation the value of A calculated in terms of m and t . After numerical substitutions one obtains the equation

m,

where m is in mg. see.-’, t in sec., Co in millimoles (9) B. 0.Pierce, “ A Short Table of Integrals,” 3rd Editibn, Ginn and Co., Boston, Mass., 1929, p. 120.

(10) D. Ilkovic, Colkclion Csechoslou. Chcm. Commun., 6, 498 (1934).

4946

per liter, k in cm. sec.-l, D in cm.2 sec.-l, and i in microamperes. @ ( k d t l D )represents the error integral ? -

Formula (6) is the general equation for polarographic limiting currents controlled by diffusion and by rate of reaction. The Ilkovic equation corresponds to the particular case of formula (6) for which the rate constant is infinite. The equation reported by Koutecky and Brdicka4 is of the same general form as equation (6). As in the case of the Ilkovic equation one could take into account the corrections proposed by Lingane and Loveridge" and by von Strehlov and von Stackelberg.12 However, these corrections will not be introduced in the present treatment. Average Limiting Current and Discussion Equation (6) expresses the variations of current i as a function of time during the life of a mercury drop. The average current is given by the equation zarernge

Vol. 73

PAUL DELAHAY

=

1255 nni2/aCok

tl/a 7b

0

[1 -

+ (k&)]

I

I

iaverage = 1255 BnmP/wg/sCOk

where function P

I

I

(8)

0 is defined by

= Y

[I

- 0 (k-\ig)]

k2r

expz

(9)

The values of 0,calculated by applying formula (9) and by using the diagram of Fig. 1, have been determined for different values of kD-'/p. The results are shown in Fig. 2 in which pkD-'/z is plotted against drop time.

exp k I t dt (7)

where r is the drop time in sec. The integral appearing in formula (7) can be calculated by expanding the error integral, but i t is much easier to determine the average current by graphic integration. The function appearing under the integral sign was plotted as a function of T for different values of kD-'l2, and the ratio y of the average value of the current to the maximum value was determined by graphic integration for different values of r . The average current is thus obtained by multiplying the second member of equation (6) by y, which is directly read on the diagram of Fig. 1. 0.9

The values of the error integral in formula (7) used in the reparation of Fig. 1were taken from tables13 for kt'fD-'/Z smaller than 1. For larger values of kt'/sD-'/Z the error integral was expanded. In the latter case the accuracy was better than 0.5%.14 Figure 1 shows that y is practically equal to 0.60 when kD-'/a is smaller than 0.05. If kD-'/z is larger than 5, y is practically equal to 0.857 or 6/, which is the value corresponding to the Ilkovic equation. In the former case the limiting current is proportional to t 2 / b , whereas it is proportional to t 1 I 8in the latter case. In order to simplify the application of formula (6) it is useful to apply the abridged equation

0.5

I

9

a 0.4

I

I

. 3 \\

1

1

G

e l

' w . 9 0.4

-.

2. II

: r

h 0.1

2

0.2

jUni1

0

2 3 4 Drop time (sec.). Fig. 2.---Variations of j3kD-1/2 as a function of drop time for different values of kD-l/Z. kD-l/2is expressed in sec.--''2. The dotted curve corresponds to @kD-'/P equal one-half its value for k infinite. 1

When the rate constant k is infinite, ;.e., when the Ilkovic equation is applicable, one has 3 3 1 Drop time (sec.). b'ig. 1. -Variations of y as a function of drop time for different values of kD-'/2. kD-'I2 is expressed in sec.-l/Z.

Figure 2 shows that equation (6) is almost identical to the Ilkovic equation for values of kD-'/%

(11) J. J. 1,ingane anql H . A . I,overidge, THISJOURNAL, 72, 438 (1950). (12) hl. \'on Strehlov and If. von Stackelberg, 2. Elrklrockein., 64, 51 (1'350).

(13) 13. Jahnke and I?. Em&, "Tables of Functions." Teubner, Leipzig, 1938, p 24. (14) T h e error is smaller than the last term of the series which is considered in the calculation.

1

I

Oct., 1951

POLAROGRAPHIC CURRENT CONTROLLED BY REACTION RATEAND DIFFUSION

larger than 5. The limiting current is thus practically controlled by diffusion. On the other hand, /3 is practically equal to 0.60 when kD-'/l is smaller than 0.05, and the current is then controlled by rate. For values of kD-'/* comprised between 0.05 and 5, both diffusion and rate determine the magnitude of the limiting current. Calculation of the Rate Constant from Experimental Data.-The rate constant k can be calculated from the experimental current i by application of formula (6) using the values of y read on Fig. 1. However, this procedure is tedious since the last three factors in equation (6) are functions of the unknown rate constant k . The following graphic procedure is much simpler and gives accurate results. The value of flk is calculated from the average value of the experimental limiting current by application of formula (8). The corresponding value of the function /3kD-'/s is calculated. The point having the drop time as abscissa and BkD-'/a as ordinate is located on the diagram of Fig. 2, and the value of kD-'/s for the curve passing through this point is determined by interpolation. The rate constant k is calculated from the value of kD-'/z. The value of the diffusion c o a c i e n t needed in this calculation may be determined by conventional methodsls or it may be calculated from polarographic data obtained with substances which may be assuined to have approximately the same diffusion coefficient as the substance being investigated. When kD-'/a is smaller than 0.05 the rate constant is directly computed from the limiting current by application of formula (8) in which fl is made equal to 0.60.

Dependence of the Limiting Current on the Head of Mercury The rate of flow of mercury m is proportional to the head of mercury H corrected for the back pressure. l6 The drop time is inversely proportional to I€. Thus one has

150

4947

1

600 800 Head of mercury (mm.). Fig. 3.-Variations of instantaneous limiting current as a function of the head of mercury for different values of kD-lI2. kD-'I2 is expressed in sec.-l/Z. Current is in per cent. of the value for H = 400 mm. 400

Ilkovic equation. When kD-'la is smaller than 0.05 the current is independent of the head of mercury. It should be pointed out that formula (13) gives the instantaneous limiting current. The average current is obtained by multiplying the second member of equation (13) by y. However, y varies with T (see Fig. l), and consequently one should take into consideration the corresponding variations of y with the head of mercury. The values of y to be used are easily determined on the diagram of Fig. 1.

Application to the Reduction of Weak Acids We consider the case of a reducible weak acid HA, and we assume that the anion A- is reduced a t more negative potentials than the undissociated acid HA. Two waves may be observed under m =m a (11) these conditions. The first wave corresponds to the TO reduction of the undissociated acid, the second to H the reduction of the anion. The limiting current of the first wave depends on the concentration of unwhere maand TO are constants. By introducing these values of m and r in equa- dissociated acid and on the rate of recombination tion (6) one obtains the instantaneous limiting cur- of the ions H + and A-, among other factors." In the present treatment we shall assume that the PH rent as a function of the head of mercury of the supporting electrolyte is adjusted to a value corresponding to an almost complete dissociation of the acid. Moreover, the potential of the mercury (13) drop is adjusted to a value a t which only the undisEquation (13) has been applied to various cases sociated acid is reduced. The number of moles of for which kD-'/* is comprised between 0.05 and 1.4, undissociated acid HA transformed a t the surface of 7 0 being equal to 1000 sec. mm. Limiting currents the dropping mercury electrode in time dt, per unit have been calculated in per cent. of the current cor- area, is d x H A = K [ H + ] [ A - ] d t - Kd[HA]cit responding to a head of mercury of 400 mm. The (14) results are shown in Fig. 3 for values of the head of where [H+], [A-] and [HA] are the concentrations mercury comprised between 400 and 800 mm. of the corresponding species in moles per ~ m . ~ . When kD-'/* is larger than 1.4 the current is prac- K and K d are the rate constants for the combination tically proportional to €2'1' as in the case of the and for the dissociation process, respectively. The (15) A. Weissberger, "Physical Methods of Organic Chemistry," second term in the second member of equation (14) 2nd edition, Vol. I, Interscience Publishers, Inc., New York, N. Y., can be neglected since we assume that [HA] is neg1941, p. 551. $ - = -

(16) See ref. (5), p. 67.

(17) See ref. 2.

PAULDELAIIAY

4948

ligible in comparison with [A-1. iMoreover, we assume that the supporting electrolyte is buffered in equation (14) is and consequently the term [H+] constant. After these simplifications equation (14) becomes similar to equation (1) and the theory developed in the present paper is applicable to the reduction of weak acids. The constant k of equation (1) is related to constant li by the equation k = lij1r+j10-3

(I51

where K is in cni. set.-' (moles per liter)-l and [ H + ] in moles per liter. It should be pointed out that k defined by equation (1) does not depend on the units in which the concentration is calculated whereas K defined by equation (14) depends on these units. Such a difference is to be expected since the rate constant k is defined for a first order reaction whereas I< corresponds to a second order reaction. pH of the Solution Exhibiting a Limiting Current Equal to One-half the Diffusion Current.When the pH of the supporting electrolyte is increased, the rate constant k defined by equation (15) decreases. The corresponding decrease in the limiting current of the undissociated acid becomes noticeable when kD-'/z is smaller than 5 (see above). Thus, there is a value of k for which the limiting current is equal to one-half the diffusion current which is obtained a t low fiH's. This value of k should satisfy the condition PkD '

9

=

0 5 X 0 4#3 -5-

(16)

4

which is derived from equation (10). By combining equation (16) with formula (9), one obtains an equation which can be solved for k. However, it is easier to determine k by the following graphic method. Values of /3kD-'/p given by equation (16) are plotted against T (dotted line in Fig. 2 ) . The point on this curve having a given drop time as abscissa is located. The curve of the family represented in Fig. 2 which passed through this point is determined and the corresponding value of kD-'/z is calculated by interpolation. The pH corresponding to i = i d / 2 is calculated by application of formula (17) which is A modified form of equation (15). (pH)]/, = log KD-'/z - log (kD- '/Z)I/~-3

(ti)

where -log(kD-'/g)l/2 is given in Table I. TABLE I 1 alues of kD-'/?and -log (kU '11) for I k o p time,

kU"

'

= ?d/2

L

- log

(LU-'/%)

sec

sec

1 1 5

0 64 51

2 0 2 S

41

3 0

38

dh 42

3 5 4 0

36 34

44 4i

47

2

u

10 27 33

Formula (17) and Table I show that the variation of ( ~ H ) I with / , drop time is by no means negligible. The variation is especially noticeable for drop times shorter than 3 secouds.

Vol. 7 3

Experimental Verification in the Case of Pyruvic Acid Experimental.-The supporting electrolytes used were Clark and Lubs biphthalate and monopotassium phosphate buffers prepared according to Kolthoff and Laitinen.18 pH's were measured with a Beckman model G instrument using a saturated solution of monopotassium taxtrate as a primary standard. l9 The concentration of pyruvic acid was calculated from the weight of dissolved acid. Waves were recorded with a Sargent polarograph model XXI. A conventional cell with a mercury pool as anode was used. The rate of flow of mercury m = 1.53 mg. sec.-l (for a head of mercury of 538 mm.) was determined with a 1.5 volt difference of potential applied on the terminals of the cell. This voltage corresponds t o a point in the upper plateau of the wave for the undissociated acid. The drop time determined under identical conditions was 3.85zec. The temperature of the solution was 28.0 =I= 0.1 . The oscillographic recording was made with a Du Mont cathode-ray instrument model 304 H. In this measurement the rate of flow of mercury was 2.00 mg. sec.-l. A direct current preamplifier was used in order to reduce the ohmic drop in the circuit to a value of the order of 0.005volt. Variation of Ldting Current with pH.-The waves obtained at various PH were essentially the same as those reported in the literat~re.*~J' Below a pH approximately equal to 4.5 one wave of constant height is observed. This wave corresponds to the reduction of the undissociated acid. Between PH 4.5 and 7 the height of this wave decreases and a second wave, due t o the reduction of the anion, appears. Above PH 7 only the wave corresponding to the reduction of the anion is observed. Limiting currents for the undissociated acid a t different H are listed in Table 11. The corresponding values of kD-I!, also given in Table 11, were determined as follows. T A R L F , 11 AVERAGELIMIUNCCURRENTS FOR UNDISSOCIATED PYRUVIC ACIDA S .4 FUNCTION OF p H

K D -1 i,

r 11

JcD-I/?,

10-0 amp.

sec.-l/2

4.20 5,20 3 . 89 5.95 6.11 6.32 ii 30 6.87

5 .i 8 5.21 3.35 2 45

... 1.7 0.43 ,26 ,083

I .04 0.79 ti3

,42

1

'2

see.

x

10-8

--'I2

(moles per liter) - 1

,060 ,045 ,031

Average

.. 2.70 :3 . 33 2.32 1.07 1.23 1.63 2.30 2.08

The limiting currents for the undissociated acid were calculated in per cent. of the diffusion current observed a t low BH, i . e . , at pH 4.2. The corresponding values of BkD-'/Z and kD-'/i were directly read on the diagram of Fig. 2 . The values of kD-'/* were calculated from kD-'/p by application of formula (15). The fluctuations in the values of KD-'Iz of Table I1 are probably caused by local variations of the hydrogen ion concentration in the immediate vicinity of the electrode. In the calculation of KD-'/r from kD-'/z one assumes that the hydrogen ion concentration a t the surface of the electrode is the same as the bulk concentration. This would be true if the dissociation or the recombination of the ions of the buffer mixture were instantaneous. Since this is certainly not the case, as it has been pointed out by Koutecky and Brdicka,2* variations of KD-I/a are observed. The variations of pH, which account for the fluctuations of (18) I. M. Kolthoff and H. A. Laitinen. "pH and Electro Titrations," John Wiley and Sons, Inc., New York, N. Y . , 1941, p. 35. (19) J. J. Lingane, A n d . Chcm., 10, 810 (1947). (20) R. Brdicka, Colbclion Csechoslow. Chem. Commun., 12, 212 (1947). (21) 0. H. Muller and J. P. Baumberger, THISJOURNAL, 61, 500 (1939). ( 0 2 ) See ref. 3.

KD-%, can be calculated from (15) on the basis of the average value of KD-l/*. The pH corresponding to the largest value of KD-lh (Table 11) is 5.68 whereas the experimental value is 5.89. The pH corresponding t o the lowest value of KD-% (Table 11)is 6.40 whereas the experimental pH is 6.11. From these data one concludes that local variations of pH a t the surface of the mercury drop are of the order of 0.3 PH unit. This figure represents only an order of magnitude because of the approximative character of calculations based on the assumption that the average value of KD-'/* of Table I1 is the correct value of KD-'/%. In further calculations we shall use the actual value of KD-lln obtained a t a given pH rather than the average value. Value of (pH)I/,.-On the basis of the extreme values of KD-'/r listed in Table I1 one obtains the followine values of (pH)l/, by application of formula (17): ( ~ H ) I / ~ - 5.49 = for KD-'/r = 1.07 X IO8 and (pH)/, = 5.98 for KD-% = 3.33 X IO8. These values are in fairly good agreement with the experimental value (pH)i/, = 5.91.Variation of the Limiting Current with the Head of Mercury.-Formula (13) was verified in the case of pyruvic acid a t pH 5.95. The actual value kD-'/l given in Table I1 was used in the calculation. The results are listed in Table I11 for different values of the head of mercury corrected for the back pressure. TABLE I11 AVERAGELIMITINGCURRENT WITH THE MERCURY FOR PYRUVIC ACIDAT pH 5.95

17ARIATIONS O F

HEADOF

5r

POLAROGRAPHIC CURRENT CONTROLLED BY REACTION KATEAND DIFFUSION

Oct., 1951

mm.

H,

Experimental average current, 10-6 amp.

Calcd. av. current. 10-8 amp.

240 364 52 1 683

2.02 2.17 2.45 2.64

2.11 2.33 2.47 2.66

The agreement between calculated and experimental values is fairly good if one takes into account the error caused by local variations of PH at the surface of the mercury drop. The difference between the values 2.45 and 2.47 microamperes for H = 521 mm. results from approximations in numerical calculations. The same value should have been obtained since the value 2.45 microamperes was used in the calculation of kD-'/n. Variation of Limiting Current with Time.-Equation (6) was verified by recording the current during the life of a drop for the solution of pyruvic acid at pH 5.95. Figure 4 shows that the agreement between the experimental curve and the calculated points is excellent. The following values were used in the calculation: m = 2.00 rng. set.-', kD-'/z = 0.26 sec.-'/z and CO= 1.50 millimoles per liter. The ratio y of the average current to the maximum current determined by graphic integration of the experimental curve is 0.65. This is in good agreement with the theoretical value 0.66 read in Fig. 1. Calculation of Conventional Rate Constant K'.-As it was pointed out in a previous section it is possible to convert the rate constant K of equation (15) into a conventional constant K', expressed in sec.-l (moles per Iiter)-l, by di-

4949

4

0

1 2 3 Time (sec.) Fig. 4.-Oscillographic recording of current during the life of a drop for pyruvic acid in buffer of pH 5.95. Dots are calculated points.

viding K by the quantity 6 (see formula 2). As a first approximation one may consider 6 as equal to the average distance of two molecules of pyruvic acid in solution. In the present case one calculates 6 = 10-6 cm. on the basis of the concentration CO = 1.5 millimoles per liter. The diffusion coefficient for pyruvic acid calculated by application of the Ilkovic equation t o the diffusion current measured a t low pH was 0.36 X 10-6 cm.4 sec.-l. Using these data and the values of KD-l/z of Table I1 one obtains the extreme values 2.0 X lo1] and 6.0 X lO11sec.-l (moles per liter)-l for K'. It should be pointed out that only the order of magnitude of K' can be obtained because of the uncertainty of the value of 6.

Conclusion The theory reported in the present paper accounts for the characteristics of polarographic currents controlled by rates of reaction and by diffusion. The Ilkovic equation corresponds to a special case of a more general formula taking into account rate and diffusion effects. This theory and its implications can be verified to a certain extent by studying the reduction of a weak acid such as pyruvic acid. Acknowledgment.-It is a pleasure to acknowledge the help of Dr. G. JaffC: in the treatment of the linear diffusion problem. BATONROUGE,LA.

RECEIVED DECEMBER 27, 1950